A source generator based modeling framework for synthesis of speech under stress | IEEE Conference Publication | IEEE Xplore